# [NEW] Anyscale Agent Skills: Optimize Ray Deployments

Canonical: https://brew.new/browse/templates/email/pt1_k97renr3hp2ctqmwx95yjpn9ns8e58yv

Brand: anyscale.com
Category: newsletter

![Preview of [NEW] Anyscale Agent Skills: Optimize Ray Deployments](https://cdn.brew.new/email-preview-426f03c53e686fe9-tracking_r57r642aswfrrd5jqhsx55e5858dt5kg-1788999158036.png)

## Email content

Hi Thomas,

Today we're announcing Anyscale Agent Skills, generally available for all Anyscale customers.

AI coding agents like Claude Code and Cursor are now how developers build and ship software. But for Ray workloads, they break down in the same ways every time: wrong GPU configs, stale APIs, broken service files, and no way to debug a live cluster failure. Every gap costs GPU hours and delays production.

Agent Skills encodes the expertise of the team that built Ray directly into your coding agent. Three skill sets are available today via the Anyscale CLI:

Workload Skills - generate code and configs for LLM serving, distributed training, batch inference, and embeddings

Platform Skills - full debug-fix-deploy loop for live Ray workloads, inspect logs, patch code, redeploy in one conversation

Infra Skills - guided end-to-end Anyscale deployment on Kubernetes or cloud VMs

Install in seconds via the Anyscale CLI:

Screenshot 2026-04-21 at 2.23.39 PM

In parallel, Anyscale is launching an Optimization Services Program where AI agents paired with our forward deployed engineers analyze throughput bottlenecks and GPU waste to generate tuning recommendations to help optimize cost and performance of production AI workloads.

Read the full blog: https://anyscale.com/blog/announcing-anyscale-agent-skills-ray

Request Access Here

Anyscale_logo_grey_blue

Advance your AI Platform with Anyscale.

LinkedIn

X

Facebook

Anyscale, Inc., 600 Harrison St, San Francisco, CA 94107, USA

Unsubscribe

Manage preferences

[Open and remix this design](https://brew.new/browse/templates/email/pt1_k97renr3hp2ctqmwx95yjpn9ns8e58yv)

[Browse email designs](https://brew.new/browse/templates)